An Open and Reproducible Deep Research Agent for Long-Form Question Answering

Ikuya Yamada, Wataru Ikeda, Ko Yoshida, Mengyu Ye, Hinata Sugimoto +3 more
12/15/2025
cs.CL

Official code: efficient-deep-research/efficient-deep-research (5 stars, Apache-2.0, last commit 2025-12-04).

Abstract

We present an open deep research system for long-form question answering, selected as a winning system in the text-to-text track of the MMU-RAG competition at NeurIPS 2025. The system combines an open-source large language model (LLM) with an open web search API to perform iterative retrieval, reasoning, and synthesis in real-world open-domain settings. To enhance reasoning quality, we apply preference tuning based on LLM-as-a-judge feedback that evaluates multiple aspects, including clarity, insightfulness, and factuality. Our experimental results show that the proposed method consistently improves answer quality across all three aspects. Our source code is publicly available at https://github.com/efficient-deep-research/efficient-deep-research.

View on arXivView PDF

Code Implementations(1)

efficient-deep-research/efficient-deep-researchOfficial100%
Apache-2.0

NeurIPS 2025 MMU-RAG Submission: An Open and Reproducible Deep Research Agent for Long-Form Question Answering

51Python, DockerfileAug 27, 202510 months agoApache-2.0
deep-researchdeep-research-agentneuripsneurips-2025retrieval-augmented-generation

Similar Papers

Harnessing Multi-Role Capabilities of Large Language Models for Open-Domain Question Answering

86% match1 repo152024

Deep Research: A Systematic Survey

84% match1 repo3182025

Long-form Question Answering: An Iterative Planning-Retrieval-Generation Approach

83% match2023

ManuSearch: Democratizing Deep Search in Large Language Models with a Transparent and Open Multi-Agent Framework

83% match1 repo352025

RAG-QA Arena: Evaluating Domain Robustness for Long-form Retrieval Augmented Question Answering

83% match1 repo532024

Comparing the Performance of LLMs in RAG-based Question-Answering: A Case Study in Computer Science Literature

83% match2025

Cite this paper

@article{yamada2025open,
  title  = {An Open and Reproducible Deep Research Agent for Long-Form Question Answering},
  author = {Ikuya Yamada and Wataru Ikeda and Ko Yoshida and Mengyu Ye and Hinata Sugimoto and Masatoshi Suzuki and Hisanori Ozaki and Jun Suzuki},
  year   = {2025},
  eprint = {2512.13059},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2512.13059v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2512.13059.svg)](https://codeofpaper.com/paper/2512.13059)

Discussion